Introduction

The introduction of PACT (Prescription Analysis and Cost) data, primary care drug budgets, and professional prescribing advisers has brought with it a whole new dictionary of words, abbreviations, and acronyms that has allowed the privileged few to talk a new and often bewildering language, whilst confusing the rest of the population. This language has become a part of general usage and an understanding is important for all of those (e.g. GPs, nurse prescribers, practice managers, pharmacists, managers, and academics) who are involved in prescribing.

The purpose of this web page is to provide a brief explanation of prescribing measures and their application for those who are new to this area. Where applicable, a brief history and relevant references are also included. This page replicates the booklet "Prescribing Measures and their application - An explanation".

Traditionally, prescribing has been measured using the number of items prescribed and the cost of these drugs both to compare health authorities and practices, and also to look at historic trends.

This report will define the historic measures currently used in reports produced by the Prescription Pricing Authority (PPA), outline new measures that are now available, and discuss how these measures can be used to produce prescribing performance indicators. The use of these indicators will be vital in the continued performance and financial management of the ever increasing drugs bill.

1. Traditional Prescribing Measures

• 1.1 Drug costs

Costs of drugs are expressed in two ways by the Prescription Pricing Authority (PPA).

• Net Ingredient Cost (NIC)

The net ingredient cost is the basic price of a drug i.e. the price listed in the Drug Tariff or MIMS. It is used in PACT reports and other analyses reflecting the prescribing performance of GPs and health authorities.

• Cash

Cash is closer to the true price paid by the NHS. Community pharmacists are reimbursed for prescriptions they have dispensed on the basis of the NIC less the discount that they are assumed to have received from their suppliers. A container allowance is then added. The discount assumed for individual pharmacies varies according to the number of items they dispense and is determined nationally by the Secretary of State for Health. However, for the purposes of GP and health authority prescribing budgets and statements a national discount percentage is used based on the total discount received by pharmacies.

Until September 1996 drug budget allocations were made to fundholding GPs in cash and to non- fundholders as NIC. All budget allocations are now made in cash. This change has been made to simplify the current system and to ensure that GPs prescribing decisions more accurately reflect the true cost to the NHS and enable more meaningful comparisons between all practices.

It is obviously vital to measure the costs of prescribing in order to financially manage the current drugs bill, and to provide forecasting models that will aid risk management while allowing for the prioritisation and introduction of new drugs.

Cost is also an important measure to use when comparing practices and health authorities, and has the benefit, when using NIC, of being standardised throughout GP prescribing in England. The NIC of any prescribed drug is the same whichever GP prescribes them.

National and international comparisons based on cost can often be misleading in the evaluation of drug use. Price differences between alternative preparations and different national cost levels make the evaluation difficult. Long term studies are also difficult due to fluctuations in currency and changes in price. When cost data is used, a significant increase in the use of cheaper drugs may have little influence on the total level, while a shift to more expensive drugs is more readily noticed.

• 1.2 Volume

Common physical units (e.g. grams, litres), number of tablets, and number of items are used for quantifying drug usage. These units can be applied only when the use of one drug, or of specific products, is being evaluated. Problems arise when the utilization of whole drug groups is considered.

If usage is given in terms of grams of active ingredients, drugs with low potency will have a larger fraction of the total than drugs with high potency.

Counting numbers of tablets also has disadvantages because strengths of tablets vary, with the result that low strength preparations contribute relatively more than high strength preparations. Also short-acting products will contribute more than long-acting preparations.

Numbers of items do not give a good indication of total use unless total amounts of drugs per item are also being considered. Counting of items however is of great value in measuring the frequency of prescribing and this is relevant to treatments given largely or entirely as courses, e.g. antibiotics and immunisations.

• 2.1 Prescribed Daily Dose

A prescribed daily dose (PDD) for a drug can be determined from prescriptions and medical or pharmacy records. The PDD is the average daily amount of the drug that is actually prescribed by a specified group of prescribers for a given time period. For drugs where the recommended dose differs from one indication to another, it is important that diagnosis is linked to the PDD. Pharmacoepidemiological information (e.g. sex, age) is also important in order to interpret a PDD. The PDD is difficult to determine using PPA data alone as the duration of each item is not available. Therefore, in order to obtain a PDD for English prescribers it is necessary to use data from sample databases, e.g. IMS or GPRD, that contain this information.

• 2.2 Defined Daily Doses (DDDs)

For many purposes it is clear that measuring the volume of prescribed drugs in terms of number of items is unsatisfactory. The main problem is that a single item can be of any quantity or for any duration, e.g. 6 months or 1 week. There is a need for a system which more reliably measures drug volume.

The system of Defined Daily Doses (DDDs), developed and maintained by the World Health Organisation (WHO) attempts to overcome these problems. In this system each drug is given a value, within its recognised dosage range, that represents the assumed average maintenance dose per day for a drug used on its main indication in adults. It is emphasised that the DDD is a unit of measurement; it is not a recommended dose and may not be a real dose. Using this system the amount of an individual drug can be expressed in DDDs and, since the DDD of one drug is assumed to be functionally equivalent to the DDD of any other drug used for a similar purpose, the number of DDDs for two or more such drugs can be added together. It is also possible to add together the DDDs of all the drugs in the same broad therapeutic class or of all the drugs given to one or more patients. By extension, cost per DDD across groups of drugs ( a measure of economy) may be compared between practices, health authorities, and regions over time indicating where higher cost alternatives have been used.

Example

Section 10.1.1 of the BNF is non-steroidal anti-inflammatory drugs (NSAIDs). The following table shows the WHO DDDs for NSAIDs and shows how the total use of these drugs within a practice can be calculated.

*calculated from the sub-total chemical substance part of the PACT catalogue

There are particular groups where the concept of a DDD is inappropriate.

• Skin preparations e.g. ointments and creams.
  The unit of issue is a tube. Patients use different quantities depending on the area to be covered and the quantity they apply per unit area, therefore it is not possible to produce a meaningful DDD.
• Vaccinations and other one-off treatments.
  DDDs are defined as the assumed average maintenance dose per day for its main indication in adults. Vaccines are not used as maintenance doses.
• Combination preparations, mixtures and compounds.
  Where a product contains more than one drug it is difficult to determine for which component the DDD should be used.
• Contraceptive pills and hormone replacement therapy (HRT) regimes.
  For HRT and contraceptives it is not possible to have a DDD. Different preparations are given for varying time periods within the month, e.g. an HRT patch may be applied twice weekly and an HRT tablet taken daily, some products are taken continuously whilst others are taken for 3 weeks of each month. Other products contain mixtures of tablets and patches. Oral contraceptive prescribing is by packs rather than tablets.

DDDs have been defined by the WHO based on international prescribing habits. Work done by the PSU has demonstrated that the prescribing of GPs in England can differ from the international standard. Reasons for variation include, for example,

• different indications for drugs used in England, e.g. the DDD for quinine is based on the dose used for malaria prophylaxis (1200mg) but in England the main indication is the treatment of leg cramps (300mg).
• prescribing habits of GPs in England, e.g. the DDD for dothiepin is 150mg which reflects the treatment dose for tricyclic antidepressants. In England GPs tend to use smaller doses of tricyclic antidepressants, and 75mg is the average dose prescribed¹.

To allow comparison of prescribing within England there is a need to have a system which more accurately reflects GPs prescribing. Hence Average Daily Quantities (ADQs) have been developed by an expert group convened by the PSU. See also: Members of this group

The following information is considered when defining an Average Daily Quantity;

1. the Defined Daily Dose (if one is available), as the World Health Organization Advisory Group have much experience in this area and have access to a variety of data sources when defining values. However, it should be noted that DDDs are an international compromise and do not necessarily accurately reflect prescribing patterns in England.
2. the Prescribed Daily Dose, if available and when calculated on a large enough sample of items, should also be considered, as it reflects the actual usage by GPs. However, it may well be that the single value Prescribed Daily Dose hides a wide variation in prescribing practice, again stressing the nature of the Prescribed Daily Dose and the subsequent ADQs as being analytical units.
3. PPA data, which gives the number of items prescribed by particular quantities of each drug preparation. This information source has the advantage of being based on every prescription dispensed in England but the disadvantage of not including the intended duration for the item, making the calculation of an accurate Prescribed Daily Dose impossible.
4. BNF information regarding dosage, particularly for maintenance doses.
5. whenever possible, therapeutic equivalence between drugs of the same therapeutic type is sought. However, where there is a discrepancy between actual usage as suggested by data sources (ii), (iii), and (iv) and equivalence data from clinical research, then the actual usage is given priority. The expert group stresses that these discrepancies should be kept to a minimum and that when they occur, should be noted in any disseminated information regarding the ADQs.
6. following the first five principles, an ADQ is set only with the agreement of all members of the group.

ADQs will be reviewed on a regular basis, thus reflecting any changes in drug utilization and the introduction of new drugs.

To date, ADQs have been developed for the following

These sections account for the majority of commonly prescribed items in general practice.

(Copies of ADQ values are available from the PSU.)

For these groups, approximately 70% of the ADQ values are the same as the existing DDDs.

The PPA have committed themselves to populating their database with ADQs. How the data will be presented in PPA information systems is currently being discussed.

3. Patient Denominators

It is obvious that the volume and cost of prescribing is influenced by the size of the population served. Practices and health authorities have therefore been compared using the ratios items per patient, and cost per patient.

General practitioners have received reports containing data about their prescribing for many years, and these have always included comparisons with local and national averages. In an attempt to make the comparisons more valid, a weighting factor, called the prescribing unit (PU), was introduced in England in 1983 to take account of the greater need of elderly patients for medication. Each patient aged 65 and over counted as three prescribing units whilst those aged under 65 counted as one. In 1993 a more sophisticated weighting system using ASTRO-PUs² was adopted which took account of sex and temporary resident status, as well as a greater number of age bands. More recently, in 1995 a system of weighted prescribing units within therapeutic groups (STAR-PUs) was devised².

PUs were adopted to take account of the greater need of elderly patients for medication in reporting prescribing performance at both the practice and health authority level. Rather than compare the cost of prescribing or the number of items prescribed by patient, comparisons by PU would weight the result according to the number of elderly patients in either the practice or health authority. A large difference in cost per PU between two practices could not then be explained by one practice having a very elderly list population.

Currently prescribing units are used by the PPA as the weighting factor in the PACT Standard Reports. Patients aged 65 and over are counted as 3 prescribing units and patients under 65 and temporary residents are counted as 1.

Prescribing units are used within the PACT Standard Reports in calculating differences in practice prescribing from the health authority (HA) average.

To calculate the HA spend on a therapeutic area, for example, the total spend within the HA is divided by the total number of PUs within the HA and then multiplied by the number of practice PUs.

Derived by the Prescribing Research Unit (PRU) in 1993², ASTRO-PUs were designed to weight individual practice populations for age, sex and temporary residents.

ASTRO-PU figures are based on cost, rather than number of prescription items (see box). They have subsequently been used by the NHS Executive and health authorities in prescribing allocation methodology and as the appropriate denominator when comparing the costs of prescribing between practices or between HAs.

Prescribing habits change over time; new drugs are introduced, older drugs may become less popular, new indications are licensed for existing drugs and more patients may be treated for existing indications as better evidence for efficacy is demonstrated e.g. the use of statins for primary and secondary prevention of coronary heart disease. For these reasons the ASTRO-PU values have been revised following work carried out by the PSU³ and to avoid confusion the new values are known as ASTRO(97)-PUs.

3Similarly for Dr. Lectors practice:

The following graph shows the difference in NIC/patient and NIC/ASTRO(97)-PU (re-scaled) for individual practices within a health authority. This demonstrates how misleading NIC per patient can be. Several practices with low and medium NIC per patient may be overlooked in an exercise to find which practices were more expensive for a comparable population. If the age and structure of the populations is taken into consideration by using NIC per ASTRO(97)-PU, their costs are somewhat higher and would draw more attention than their NIC per patient. NIC per patient clearly hides the high cost of this practice!

ASTRO-PUs are devised from the total of all drug costs. It is therefore not strictly correct to use NIC per ASTRO-PU for making comparisons within a therapeutic group e.g. NIC/ASTRO-PU for ulcer-healing drugs. There are differences in the age and sex of patients for whom drugs in specific therapeutic groups are usually prescribed. To make such comparisons STAR-PUs⁴ have been developed along similar lines to the ASTRO-PU but based on costs within therapeutic groups. These were recently revised as STAR(97)-PUs.

STAR(97)-PUs have been developed for the 8 leading therapeutic groups, which together account for 85% of prescribing in England. These therapeutic groups and the corresponding STAR(97)-PU values are detailed below. The weightings are based on cost rather than number of prescription items.

STAR(97)-PU values

STAR(97)-PU values are now available for the following BNF sections to comply with some of the groups for which ADQs have been developed. Eventually STAR(97)-PU values will be available for all sections for which ADQs have been developed.

4.1 Prescribing Indicators

Prescribing indicators are measurements that are used to show how prescribing performance for a health authority or a practice might compare with other bodies, national averages or themselves over time. Prescribing indicators should not be seen as definitive or be used in isolation but rather as a starting point for reviewing progress.

A Prescribing Indicator Group (PIG) with membership from various parts of the NHS and academia examined the development and maintenance of a core set of such indicators. Suggested indicators need to be simple in concept and based on justifiable criteria. The criteria used by PIG recommends that indicators should be;

• based on clinical evidence or professional consensus.
  Prescribing indicators should encourage good prescribing which is also cost effective. As well as measuring prescribing, any indicator should aim to change behaviour.
• accepted as relevant and useful
  Prescribing indicators should be acceptable to those to whose behaviour and practices they may be applied.
• based on reliable, accurate and comparable data.
  Within the NHS most prescribing indicator data will be derived from PACT. Patients may be consuming medication which is not dispensed on an FP10 e.g. aspirin 75mg bought OTC, or antibiotics obtained on a private prescription.
• able to demonstrate changes in prescribing behaviour.
  Changes in prescribing behaviour for drugs which are for short courses e.g. antibiotics will be demonstrated by prescribing indicators earlier than changes in prescribing of drugs used for chronic conditions and which are subject to repeat prescribing e.g. acid suppressing therapy. If the indicator does not measure prescribing over a long enough period, or is not repeated frequently enough, then changes in prescribing behaviour which revert back to the previous pattern will not be identified.
• appropriately weighted to allow comparisons between practices or HAs and to allow changes over time.
  Variables within a population need to be controlled to make relevant comparisons e.g. age and sex. Appropriate units of measurement should be used.
• able to discriminate between more and less desirable prescribing behaviour.
  Prescribing indicators should not be misleading.

The MEMPHIS software was developed to include a wide range of indicators to provide senior NHS management with an overview of HA performance. A limited number of performance indicators relating to prescribing were included. These indicators were intended to be simple in concept, whilst being based on professionally justifiable criteria.

Example

Drugs of Limited Clinical Value as defined by the Audit Commission

NIC/PU for Drugs of Limited Clinical Value for an individual practice

Number of PUs (patients under 65 = 1 unit, over 65 = 3 units)

This figure is not comparable with the example on the previous page as it is based on a single quarter.

Eight further indicators will be included as MEMPHIS indicators in 1998:

• DDDs per STAR-PU for oral NSAIDs
• NIC per DDD for oral NSAIDs.
• DDDs per STAR-PU for ulcer-healing drugs.
• NIC per DDD for ulcer-healing drugs.
• Items per STAR-PU for antibacterials.
• NIC per item for antibacterials.
• Number of months of treatment of HRT per woman 45-64.
• NIC per month of treatment of HRT.

5.1 ASTRO-PU Weightings for Patients in Homes

It has been suggested by Brian Crompton⁶ and others that patients in nursing homes and residential homes need higher weightings within the ASTRO-PU system because of their greater need compared with patients of the same age living in their own homes. The PSU carried out a study in collaboration with over 120 practices from 32 HAs in April, May and June of 1997 looking at the needs of patients in homes^7,8.

The ASTRO-PU weightings for patients in residential homes were recommended as being twice the value of the corresponding patients in their own homes. For patients in nursing homes, the ASTRO-PU value is three times that for corresponding patients in their own homes.

To allow for this in budget allocations for practices it is recommended that authorities do the following:

5.2 Low Income Scheme Index

The Low Income Scheme Index (LISI) is a measure of deprivation based on claims for exemption from the prescription charge on the grounds of low income⁹. The index is calculated and distributed by the Prescribing Support Unit in Leeds. The figures are collected as part of a 5% sample of prescriptions processed by the PPA. Because the figures are based on a small sample the index is not calculated for practices with less than 1,000 patients or where less than 3 items per patient are prescribed per year. Since dispensing patients usually do not receive the prescription form they cannot make a declaration and so the coverage is much less complete for doctors with large numbers of dispensing patients. The index is not calculated for practices where dispensing patients form more than one third of the practices list. The index is available to any authority who wishes to request it for their practices. At the time of writing the most recent year for which the index is available is the financial year 1995/96 but the index for the year 1997/98 should be available in June 1998.

5.3 Future Initiatives

A toolkit approach to providing prescribing information has been developed by the PSU, NPC, and PPA where HA advisers are sent standard reports relevant to policy initiatives such as performance management, resource allocation, and incentive schemes. These information sets are seen as being complementary to the various other paper reports, and query based electronic information systems provided by the PPA, not as an alternative.

The reports currently provided to HA advisers are as follows:

1. the original MEMPHIS indicators comparing all HAs with each other and the England average.
2. HA and practice data showing expenditure on a defined group of specialist drugs.
3. HA and practice level data on potential generic savings for targeted generic substitution.

It is intended to increase the number of prescribing indicators available using the ADQ drug groupings described earlier, and also to supply them at the individual practice level so that HA advisers can target practices showing extreme behaviour. The PSU is now working with the PPA to provide an electronic version of the Toolkit available alongside and complementary to the PPA EPACT query based system.

We hope you have found this report informative. If you have any queries or suggestions
for future editions, please do not hesitate to contact the Prescribing Support Unit.

PRESCRIBING SUPPORT UNIT
Brunswick Court
Bridge Street
Leeds
LS2 7RJ
Tel: (0113) 247 0777
Fax: (0113) 247 1666

E-mail: (SMTP)

or

c=gb; a=nhs; p=nhs nationalms; o=nhs prescribing support unit; s=psu (X400)

1. Donoghue J, Tylee A, Wildgust H. Cross sectional antidepressant prescribing in general practice in the United Kingdom, 1993-5. BMJ 1996;313:861-2

2. Roberts SJ, Harris CM. Age, sex, temporary resident originated prescribing units (ASTRO-PUs):new weightings for analysing prescribing of general practices in England. BMJ 1993;307:485-8.

3. Lloyd DCEF, Roberts DJ, Sleator D. Revision of the weights for the Age Sex Temporary Resident Originated Prescribing Unit. British Journal of Medical Economics 1997;11:81-5

4. Lloyd DCEF, Harris CM, Roberts DJ. Specific therapeutic group age-sex related prescribing units (STAR-PUs):weightings for analysing general practices prescribing in England. BMJ 1995;311:991-4

5. A Prescription for Improvement. Report of the Audit Commission. HMSO 1994

6. Crompton B. GPs need more PUs for nursing home patients. Prescriber 1995;1:44-5

7. Lloyd D, Scrivener G. Prescribing for patients in homes: a revised ASTRO-PU. Prescriber 1998:9:33-6

8. Petty D, Scrivener G. Prescribing patterns for patients in homes. Prescriber 1998:9:103-6

9. Lloyd DCEF, Harris CM, Clucas DW. Low income scheme index: a new deprivation scale based on prescribing in general practice. BMJ 1995;310:165-5.

Mrs. Christine Dalton - Director of Pharmaceutical Advisory Services, PPA

Dr. John Ferguson Medical Director, PPA

Professor Conrad Harris - Professor of General Practice, University of Leeds

Dr. Andrew Herxheimer - formerly of the UK Cochrane Centre

Mr. Martin Jenkins Deputy Director of Pharmaceutical Advisory Services, PPA

Ms. Margaret Maxwell - Research Fellow, University of Edinburgh

Mr. Duncan Petty - Pharmaceutical Adviser, Leeds Health Authority

Mr. Dinesh Mehta - Executive Editor, BNF

Mr. Dave Roberts - Manager, Prescribing Support Unit

Dr. David Sleator - Head of Clinical Development, West Surrey Health Authority

Professor Tom Walley - Professor of Clinical Pharmacology, University of Liverpool

Mr. John Wilson - Regional Pharmaceutical Prescribing Adviser, Trent Health Authority

ADQ Average Daily Quantity

ASTRO-PU Age Sex Temporary Resident Originated Prescribing Unit

ATC Anatomical Therapeutic Chemical, coding system for drugs

BNF British National Formulary

DDD Defined Daily Dose

Drug Tariff publication listing prices payable to pharmacists for medicines and appliances prescribable on the NHS. It also lists substances which are not prescribable, only prescribable for certain conditions and the necessary endorsements.

EPACT PPA software to submit data requests and analyse resulting data

FHSA Family Health Service Authorities (there were 90 of them in England), formerly known as FPC (Family Practitioner Committees); as from April 1996 they were merged with the DHAs to form 100 HAs

FP10 form to order drugs on the NHS. Each FP10 sheet from a prescription pad bears ID code of doctor.

GPRD General Practice Research Database formerly VAMP, now operated by ONS.

HA Health Authority, there are 100 in England responsible for primary and secondary care

HCHS Hospital and Community Health Services

IMS Intercontinental Medical Statistics Ltd, produce software for use by pharmacists and collects medical data; it has the advantage over GPRD that they can price their data but the disadvantage that they have fewer practices

IPA Indicative Prescribing Amounts, notional budgets set by the HA for practices which are non-fundholders

Jarman index An index based on Census variables intended to measure GP workload although also used to measure deprivation. Practices can receive payments based on how many patients are classified as deprived on the Jarman index.

Körner age bands age bands widely used in secondary care and defined as 0-4, 5-16, 17-24, 25-34, 35-44, 45-54, 55-64, 65-74, 75-84, 85+

LIS Low Income Scheme, operated by PPA, allows those on a low income to claim full or partial exemption from the prescription fee

LLSI Limiting Long Standing Illness, variable introduced in 1991 Census

MAAG Medical Audit Advisory Group

NIC Net Ingredient Cost

NPC National Prescribing Centre, established April 1996

ONS Office for National Statistics

OPCS Office of Population Censuses and Surveys (now ONS following merger with Central Statistical Office))

PACT Prescribing Analyses and CosT

PDD Prescribed Daily Dosage.

POM Prescription Only Medicine, i.e. a drug only obtainable on prescription

PPA Prescription Pricing Authority

PU Prescribing Unit; each person below 65 is one PU, those over 65 are 3

VAMP company which originally gave free hardware and software to GPs to help them run their practice in return for the data which it then sold. In late 1993 the company was taken over by Reuters and the data is now held by the government (see GPRD)

If you have any queries please telephone the PPA Helpdesk on
0191 203 5050